Adverse Events After Posterior Lumbar Fusion Are Not Sufficiently Characterized With 30-Day Follow-Up

A Database Study

Alexander J. Kammien, BS; Justin R. Zhu, BS; Stephen M. Gillinov, AB; Michael J. Gouzoulis, BS; Jonathan N. Grauer, MD


J Am Acad Orthop Surg. 2022;30(11):528-533. 

In This Article

Abstract and Introduction


Introduction: Many studies track outcomes after procedures, such as posterior lumbar fusion (PLF), for only 30 days because of database limitations. However, adverse events may not have plateaued by this time. Thus, this study used an alternate database to evaluate the timing of adverse events for 90 days after PLF.

Methods: Adult PLF patients were identified from the 2010 to 2020 Q2 M53Ortho PearlDiver administrative data set. Ninety-day rates of multiple adverse events were determined. The time of diagnosis for each event in the 90-day postoperative period was determined. Data were dichotomized by occurrence in days 0 to 30 and 31 to 90. Median, interquartile range, and middle 80% for the time of diagnosis were determined for each adverse event.

Results: Of 51,915 patients undergoing PLF, 7,141 (13.8%) had an adverse event within 90 days of PLF. Of these, 5,174 (72.5%) experienced an event within 30 days and 2,544 (35.6%) after 30 days. For individual adverse outcomes studied, the percent that occurred 31 to 90 days after surgery ranged from 9% to 42%. The time of diagnosis (median; interquartile range; middle 80%) for each adverse event was as follows: transfusion (2 days; 2 to 5 days; 1 to 26 days), acute kidney injury (9; 2 to 29; 1 to 60), hematoma (9; 4 to 20; 3 to 39), cardiac event (11; 3 to 43; 1 to 71), pneumonia (12; 4 to 38; 2 to 68), venous thromboembolism (15; 6 to 33; 3 to 62), sepsis (19; 9 to 39; 4 to 63), surgical site infection (21; 14 to 34; 8 to 48), urinary tract infection (22; 8 to 49; 4 to 72), and wound dehiscence (27; 17 to 39; 9 to 54).

Discussion: This study highlights the importance of looking past the 30-day mark for adverse events after PLF because approximately one-third of adverse events in this study were diagnosed 31 to 90 days after surgery. This can affect research studies, patient counseling about the incidence of specific adverse events, and the development of mechanisms for surveillance at key time points.


Posterior lumbar fusion (PLF) is a common surgical procedure for which there is notable attention in the literature focused on understanding postoperative adverse outcomes.[1–4] The advent of database research has facilitated studies of such adverse outcomes in large patient cohorts, but with different data sets having varied limitations, some studies have been limited to 30-day postoperative tracking.[5–10] This study was developed to evaluate the sufficiency of a 30-day follow-up for tracking adverse events after PLF.

Bohl et al[5] assessed the timing of adverse events after anterior cervical decompression and fusion and PLF. Because that study was done in the American College of Surgeons National Surgical Quality Improvement Program, the data elements in that study were of high quality, but it is an example of a study that was limited to 30 days of follow-up. Notably, that study found that the rates of occurrence for a number of adverse events had not plateaued by the end of the study period.

If adverse events are not studied until their rates plateau, there is potential for incompletely characterizing the postoperative events after surgeries, such as PLF. This study sought to use an alternate national database to characterize the timing of postoperative adverse outcomes for 90 days after PLF, with a specific focus on highlighting differences between findings within 30 days and between 30 and 90 days.